Investigating the Effect of Cerebral Atrophy and Age-Related Changes in Brain Stiffness on Brain Deformation

Introduction: : Traumatic brain injury (TBI) is a significant public health concern, with approximately 1.7 million cases annually in the US. Age is a critical risk factor, with adults over 65 accounting for more than 60% of TBI-related hospitalizations and deaths (Thompson et al., 2006). Despite this, the biomechanical mechanisms underlying age-related differences in injury risk remain poorly understood. Structural brain changes with aging—such as cerebral atrophy and ventricular enlargement (Alshareef et al., 2021)—are believed to alter mechanical responses to head impacts. Prior work has considered brain atrophy, but the directional nature of ventricular expansion remains unquantified, and its influence on brain deformation response is unknown.

In addition to anatomical changes, brain stiffness decreases after age 40 (Hiscox et al., 2021) due to widespread cellular degeneration. These changes alter tissue composition and affect how the brain responds to mechanical forces. Age-related degradation in material properties may significantly impact strain responses and injury susceptibility. However, the relationship between brain softening and biomechanical strain during head impacts remains unstudied.

This study uses subject-specific finite element (FE) simulations of the aging brain to (1) quantify the effect of directional ventricular expansion on brain strain during various head rotations, and (2) investigate how age-related softening influences global strain patterns. We hypothesize that both ventricular expansion direction and tissue softening affect strain magnitude during head rotation. Longitudinal magnetic resonance imaging (MRI) data from healthy older adults are used to construct subject-specific FE models that reflect individual atrophy trajectories and simulate head impacts to analyze strain responses.

Materials and

Methods: : Forty subjects with the greatest longitudinal change in normalized ventricular volume were selected from the Baltimore Longitudinal Study of Aging (BLSA), each with at least three MRI scans (Fig. 1B). This metric provides an atrophy measure independent of overall brain size. Anatomical segmentation was performed using the SLANT-CRUISE deep learning pipeline (Huo et al., 2019). Voxel-based 3D FE models were generated using a previously validated MATLAB-based meshing framework at a spatial resolution of 1.5 mm (Fig. 1A) (Alshareef et al., 2021).

Directional ventricular expansion was quantified using ventricle-to-brain ratios (VBRs), calculated by measuring maximal lengths along the x, y, and z axes. FE models were simulated with axial, sagittal, and coronal head rotations in LS-Dyna, using a half-sine angular velocity pulse with peak angular velocity and acceleration of 40 rad/s and 3600 rad/s², respectively. The 95th percentile of maximum principal strain (MPS) across the whole brain was computed to evaluate peak brain deformation relative to baseline. Direction-specific strain patterns were compared across simulations (Fig. 1C).

To investigate the biomechanical response of the softening brain, aging-related changes in material properties were incorporated into the FE models. For two representative subjects, simulations were repeated across all three rotational directions, with properties adjusted per scan to reflect an annual 0.8% decrease in stiffness, as reported by Sack et al. (2009). Strain outcomes were then compared across timepoints.

Strain outcomes were correlated with anatomical metrics to assess the combined effects of ventricular geometry and stiffness degradation on biomechanical vulnerability.

Results, Conclusions, and Discussions:: A total of 186 axial rotation simulations were performed for 39 high-ventricle-enlargement (HVE) and 9 low-ventricle-enlargement (LVE) subjects, with an additional 54 simulations in sagittal and coronal planes. Across HVE subjects, MPS decreased by an average of 0.66% per year, compared to just 0.025% in LVE subjects—indicating a significant reduction in strain magnitude with increasing cerebral atrophy. Among all loading conditions, axial rotations consistently yielded the greatest strain reductions with age, while coronal rotations produced the lowest strain magnitudes overall. Interestingly, some sagittal simulations showed strain values comparable to or greater than those from axial loading in certain subjects (Fig. 1C), suggesting that ventricular shape and expansion direction may change the rotation direction that causes the highest brain deformations.

Notably, inter-subject variability was prominent (Fig. 1D). While some subjects showed uniform strain decreases across all directions, others exhibited direction-dependent variation. These discrepancies may relate to baseline age, head size, sex, or individual patterns of atrophy. Future work will explore these factors through regression analysis and stratified modeling. Additional planned analyses include regional and voxel-wise strain evaluation to identify focal vulnerabilities linked to specific anatomical features.

When accounting for progressive brain tissue softening, the male subject exhibited a 1.067% average increase in MPS across all loading directions, while the female subject showed a more pronounced increase of 3.222%. These results suggest that age-related declines in tissue stiffness may increase strain responses, particularly in females, and highlight the need to consider sex-specific material property trajectories in future modeling efforts. More importantly, the strain trends in the different rotational directions for the subjects with material property changes were similar to those observed in the original cohort, but with elevated overall magnitudes.

Although these models capture mechanical rather than physiological outcomes, they offer key insight into the aging brain’s evolving vulnerability to biomechanical loads. The findings underscore the importance of modeling both anatomical and material property changes in TBI risk assessment. This subject-specific FE modeling framework may serve as a foundation for personalized prediction tools aimed at preventing TBI in older adults.

Acknowledgements and/or References (Optional):: Thompson, H.J., McCormick, W.C. and Kagan, S.H. (2006), Traumatic Brain Injury in Older Adults: Epidemiology, Outcomes, and Future Implications. Journal of the American Geriatrics Society, 54: 1590-1595.

Alshareef, A., Knutsen, A. K... & Ramesh, K. T. (2021). Integrating material properties from magnetic resonance …models for the human brain. Brain Multiphysics, 2, 100038.

Hiscox, Lucy V., et al. "Aging brain mechanics: Progress and promise of magnetic resonance elastography." Neuroimage 232 (2021): 117889.

Huo, Y., Xu, Z., Xiong, Y., Aboud, K., Parvathaneni, P., Bao, S., ... & Landman, B. A. (2019). 3D whole brain segmentation using spatially localized atlas network tiles. NeuroImage, 194, 105-119

Ingolf Sack, Bernd Beierbach, Jens Wuerfel, Dieter Klatt, Uwe Hamhaber, Sebastian Papazoglou, Peter Martus, Jürgen Braun, The impact of aging and gender on brain viscoelasticity, NeuroImage, Volume 46, Issue 3, 2009, Pages 652-657